This study aims to identify Age-Related Macular Degeneration on complex data (existence of other retinal diseases in fundus images containing AMD). With the rise of artificial intelligence for automation of pattern recognition, image processing would become a viable way of aiding the detection of diseases for medical science. This study involves Image Data Gathering, application of CLAHE on the green channel for Image Preprocessing, the use of UNet for Image Segmentation, and the implementation of Resnet, InceptionV3, and CNN classification models. Upon testing, The inclusion of U-Net as an image segmentation method in the deep learning architectures showed consistent improvements in the metric scores across different models. Also, it was observed that InceptionV3 with U-Net exhibited the highest metric scores across Precision, Accuracy, and F1-score with a 98.25% accuracy. The researchers recommend adding more data since the researchers were limited to open-source datasets along with using better segmentation method that does not erode or crop image features such as a deeper model similar to MSU-Net.